Making Sense Of The Google CTR Debacle

Wednesday, February 27, 2008 By Clay | Comments (4)

I’m amazed at the logic (or lack thereof) behind the recent sell-off of Google stock amid Comscore’s report of flat YOY January paid search click-through rates. The press and blogosphere has been buzzing with stories ranging from sensational (Google Disaster: Comscore Reports Awful January Data) to simply inconsistent (Google Shares Fall 8% on Ad Fears). Many stories went as far as using Comscore’s questionable data point as an indicator to Google’s susceptibility to US market conditions. Here’s a review of what happened and my take on why we should be rolling our eyes.

What Happened?

Comscore reported January 2008 click-through rates on paid search ads remained flat in comparison to January 2007. Let me repeat that, CTRs on paid ads on Google.com remained FLAT. Not traffic declined, not revenue per click declined, not net profit declined, and certainly not EBITDA declined. Is this worth raising an eyebrow? Sure. Is this worthy of a knee-jerk reaction that results in a 4.6% decline in the stock and billions of market cap value? Absolutely not.

What Could Be Causing A Deceleration in CTR Growth?

A deceleration in CTR growth may be a signal that Google’s paid advertisements are seen as less relevant to searchers. While there are a host of factors that could cause this, I think two in particular are playing a significant role; Google’s expanded broad match and paid search marketing campaign maturation.

Before jumping into each factor, there’s also a macro element at play; the long-tail of search. Search engine users tend to evolve from using one keyword per search to using two, three, and four keyword phrases. As they get more experienced using search engines they tend to use more keyword phrases (eMarketer muscle). These longer keyword phrases represent more fragmented keyword inventory (we move from ‘flights’ to ‘cheap flights to miami’) and a longer tail of keyword volume. This evolution is playing a role in the two (potential) deceleration factors:

• Google’s Expanded Broad Match: In recent months Google has been very aggressive with their expanded broad match feature.Google will take an advertiser’s ad targeted to the keyword ‘flights’ and display it on searches for ‘trips’ – a practice that’s very difficult for advertisers to control.Google has taken upon themselves to help advertisers reach the long-tail keyword queries they may not have included in their campaigns.While this delivers additional volume it also produces less relevant ads for searchers and lower click-through rates.
• Paid Search Marketing Campaign Maturation: As search engine campaigns mature, managers expand their keyword lists to find more and more search volume (more eMarketer muscle). They tend to venture outside of keywords within their primary keyword market (‘cheap flights’) and move into keyword phrases that reach their target market but may not elicit a direct response/conversion (’luggage’).This activity, particularly for large advertisers who can afford to do it on a large scale, means less relevant ads for searchers and lower CTRs.

Some other theories require a bit more data which isn’t available. I’d love to know if…

• Paid search ads are becoming more and more commercial. With CPC’s rising, are less and less content focused advertisers being pushed out by commercial/commerce focused advertisers who can attribute a direct ROI?
• Organic results, particularly Google’s move to blend images and video into results (universal search) is affecting paid search CTRs. Could the move to blend video into paid ads change this?

In summary, there are a number of factors at play causing the (potential) deceleration. Two in particular involve the long tail of search and how advertisers capitalize on it. And while this is potentially causing a deceleration in CTR’s it’s not necessarily a bad thing for Google’s bottom line. Which leads me to…

Why You Shouldn’t Get Too Excited About Comscore’s Report

Google has many ‘levers’ they can pull to improve quality and increase revenue. While CTR on sponsored links is certainly looked at, it can easily be manipulated and I highly doubt it’s one of their KPIs.

To illustrate a lever, Google has a minimum bid and CTR threshold for an ad to qualify for the premium position (top left above the organic results). Ads that appear in this premium position receive extremely high CTR’s; well above the norm. Now, if Google was concerned about CTR (to the extent where it should be used to evaluate the health of the business and influence financial analyst buy/hold ratings) they could easily lower the threshold and allow more advertisers in that space.

Why doesn’t Google do this? Because CTR isn’t a primary KPI. Revenue (in addition to user experience metrics that encompass ‘quality’) is and with their continuous market share growth they don’t need to worry about CTR. A simple illustration:

If this was your search engine, would you like period A or period B? If A was January 2007 and B was January 2008 (yes, I accounted for the 85% YOY search query growth) would you be as concerned about Google’s health let alone their susceptibility to the market slowdown? I think not. The sell-off has created nothing but a bargain for savvy buyers.

Thoughts?

UPDATE (2/29): Comscore has released their explanation here. Very much in line with my thoughts. Choice quote…

“We generally do not comment on industry data that we release because we are expected to provide information and not opinions. However, the conclusion that was being drawn about the softening of the online advertising market, while at first glance is supported by a data sound bite like a “drop in paid clicks”, does not hold water once you dig deeper into the more detailed information provided in the paid click data. Left unchallenged, it harms the interests of the overall Internet industry. That is why we took the unusual step of writing this note.

It is important to emphasize that we are not repudiating our own data. Quite the opposite: our data remains unchanged, and, we believe, correct. We are just offering a more thorough analysis to ensure the information is interpreted correctly and that the proper conclusions are being drawn from it.”

Related reading:

• Google Watch (2/29), It’s Google, Not The Economy, Stupid
• Clickz (2/29), ComScore Weighs in on Google Click Volume Debacle Sparked by Its Data
• Barrons (2/29), ComScore: Weak Paid Clicks Reflect Google Policy Shifts
• Techcrunch (2/29), Google CTR Down Due To Click Area Changes
  
• Silicon Alley Insider, Google Disaster: Comscore Reports Awful January. Choice quote, “Yes, Comscore could be wrong, and, yes, it’s only one data point. But hard to imagine how the bulls are going to spin this one.” And the link to Google Bulls Try to Spin Comscore Disaster, Fail doesn’t seem to be working.
• New York Times, A Highflier Loses Altitude as Google’s Clicks Go Flat. A well balanced piece.
• NYT/Reuters, Google Shares Fall 8% on Ad Fears. Choice quotes, “UBS analyst Benjamin Schachter cut his 12-month target on Google from $650 to $590 — a new low among bullish Wall Street analysts”. Followed by, “Schachter maintained his long-term buy rating on Google. The UBS analyst also noted that comScore had recently revised the way it measures visits to Web sites, which made comparisons to previous trends difficult.” Huh?
• Seeking Alpha / Techcrunch, Did The Market Overreact To Google’s Paid Click Decline. The best read of the bunch. The correlation analysis from JPMorgan is particularly interesting.
• Fred Wilson, Thinking About GOOG This Morning. Always a pragmatic approach.
• Hitwise Intelligence, Google: Does Search Data Indicate a Recession? An alternative data perspective.

Comments (4) | Categories: Search Engine Industry News | Tags: comscore, google, industry news    |  Popularity: 37% [?]

The Semantic Web & Search

Friday, February 22, 2008 By Clay | Comments (0)

Bernard Lunn at ReadWriteWeb writes an excellent post on the semantic web entitled 11 Things To Know About The Semantic Web. If you’re a search marketer (paid or organic) and you’re not familiar with the emerging semantic web than you’ve got some reading to do.

Bernard touches on the decline of relational databases, the impact on Google, and what the killer app will or will not look like. Here are my thoughts on where he’s dead-on and where he may have missed it:

5. Don’t look for a killer app. That implies a client/consumer win. This is much more likely to be a server/platform/enterprise win. Even if the initial experimentation is done in the consumer domain; Freebase for example looks like a mass Beta test for some enterprise technology that Metaweb wants to release later.

The killer apps are out there. Trust me. Some are in development, some are in the wild, and some are just waiting for the enabling power of the semantic web. I know Alex Iskold would argue it’s in its infancy and well positioned for the emergence of the semantic web.

7. Semantic Web could slow the Google steamroller. This could be like the PC for IBM or the Web for Microsoft. The steamroller’s momentum carries it forward for a very long time and it can build all kinds of wrapper systems around it, but something new always does come along. Google mastered how to give some structure to countless unstructured HTML pages. Semantic Web will gradually make that less critical as the underlying content will be more structured. These big generational changes - mainframe to PC to Web - seem to be happening faster, so it seems about time for another big generational change to start happening.

Google is well positioned for the semantic web. Their algo’s are just salivating for more contextual data and their toolbar (among other tools) will evolve to provide further structure to the quasi-unstructured web. Google will face some new competition but they’re in a tremendous position for this evolution. See vertical search below.

8. But don’t look for Yet Another Search Engine (YASE) to be the David to Google’s Goliath. Just like PC was not another mainframe and Web was not another PC. Don’t ask me precisely what it will look like; if I did know I would have to kill you if I told you. I just know what it won’t look like.

I agree completely. I have a hunch what it will look like and I think it will strike Google by changing user behavior. I’ll elaborate on this when the time is right

9. Vertical Search is the pragmatist’s Semantic Web. Vertical Search businesses use whatever techniques they need - basic search engines, scrapers, APIs, human editors - to create some meaningful/useful structure in a single domain. Over time these cobbled together pragmatic solutions will be replaced by a semantic web platform, probably by an API that enables human editors to leverage their valuable domain expertise.

A fantastic point that’s often overlooked. Google looks for scalable solutions and today’s vertical search engines are very difficult to scale (across verticals). The semantic web will allow Google to build scalable apps in all kinds of verticals.

10. Tagging is the quietly disruptive technology. Everybody tags. It is the most basic human urge to mark what we find. We do it with Folders in Windows. We do it online with Bookmarks. Specialist tag Microformats such as Hcard and Hcalendar add more structure and we are only at the very start of this wave.

Another great observation. Fred Wilson pleaded with the Delicious team not to sell out and create a next generation search engine based on tagging. While Yahoo has started the integration process by displaying Delicious data in their SERPs there have been no indications the data is being used as a ranking factor.

11. Semantic Web will leverage the “community” to add structure and this will use some techniques from first generation Social Networking. But it is very unlikely that Semantic Web will emerge from the walled gardens of current social networking sites. The winners will know how to motivate community to provide structure and will provide the tools that make the structuring so easy that nobody knows they are doing anything so boring as structuring. That is the big lesson from Web 2.0 that will be applied in the Semantic Web.

Right on but an application must exist that allows the community to connect & communicate outside of their primary destination. This will come.

What are your thoughts?

Comments (0) | Categories: Semantic Web & Search | Tags: google, semantic web    |  Popularity: 27% [?]

SEO Beginners Guide from Google

Tuesday, February 19, 2008 By Clay | Comments (0)

Google has released a very useful SEO beginners guide (hat tip) entitled: Making the Most of Your Content: A Publisher’s Guide To The Web. The guide doesn’t provide any earth shattering insight (it’s for beginners) but it is a useful tool to back up your recommendations to clients/constituents. It should also help you frame your 60 second SEO basics pitch to SEO newbies.

Here’s a rundown of what’s in the guide (the index):

• A brief overview of web search
• What’s new in Google web search?
• Can Google find your site?
• Can Google index your site?
• Does your site have unique and useful content?
• Increasing visibility: best practices
• Webmaster Central
• FAQ’s
• Glossary

And here are some highlights:

1. A nice visual of Google’s process:

2. A short and concise summary of useful and unique content that’s properly optimized.

“Once the site is discoverable and indexable, the final question to ask is whether the content of the web pages is unique and useful.

First look at your text as a whole. Are your title and text links descriptive? Does your copy flow naturally and in a clear and intuitive manner?

Just as a chapter in a book is organized around specific areas and themes, so each web page should be focused on a specific area or topic. Keywords and phrases emerge naturally from this type of copy, and users are far more likely to stay on a web page that provides relevant content and links.

Make sure, however, that the phrases you write include the phrases that visitors will likely search for. For instance, if your site is for an MG enthusiast club, make sure the words ‘MG’ and ‘cars’ actually appear in the copy, rather than only terms like ‘British automobiles’.” (copyright Google, Inc.)

3. Google’s best practice recommendations:

What to do:

1. Create relevant, eye-catching content…
2. Involve users…
3. Monitor your site…
4. Aim for high-quality inbound links…
5. Provide clear text links…

What to avoid:

1. Don’t fill your page with lists of keywords
2. Don’t attempt to ‘cloak’ pages…
3. Don’t put up ‘crawler only’ pages…
4. Don’t use images to display important names, content or links…
5. Don’t create multiple copies of a page under different URLs with the intent of misleading search engines.

There are a couple of additional goodies in the guide. Give it a read.

Comments (0) | Categories: Best Practices: Organic Search Marketing | Tags: google, SEO best practices    |  Popularity: 25% [?]

Google’s ‘I’m Feeling Lucky’ Being Used For Email Spam

Friday, February 8, 2008 By Clay | Comments (0)

MSNBC’s red tape chronicles blog written by Bob Sullivan has an interesting piece on how Google search engine results pages and their ‘I’m Feeling Lucky’ feature are being used by email spammers.

“With traditional spam finally losing traction among e-mail users, spammers have stepped up their pace of innovation. Last year, they adopted new techniques like image spam, .pdf spam and even audio spam. These disappeared as quickly as they came. But starting in January, spammers began flooding inboxes with a new kind of spam that uses a much simpler form of deception. In the body of these e-mails, recipients see what looks like a link to Google search results — and in fact, that’s what it is. There’s trouble, however, on the other side of that link.

The attack combines two tactics. First, spammers game Google so the Web site they want recipients to visit ranks at the top of the search engine results. Second, they alter the URL pasted in e-mails so users who click on the link go directly to the top result via Google’s “I’m Feeling Lucky” feature – bypassing a stop at Google’s Web site.”

Apparently using a legitimate Google link bypasses spam and phishing filters. Google, always on top of their game, said they’re already working on stopping this.

“A Google spokeswoman who asked not to be named said the company has seen “I’m Feeling Lucky” attacks, but added that help is on the way.

“Google began deploying a fix that should block most of these ‘I’m Feeling Lucky’ redirects, and we will work to reduce such issues in the future,” she wrote in an e-mail.”

Full story, New Cyber Trick: Search Engine Spam.

Comments (0) | Categories: Search Engine Industry News | Tags: google, spam    |  Popularity: 20% [?]

Inside The Numbers: January Search Marketing Metrics

Thursday, February 7, 2008 By Clay | Comments (0)

eMarketer has released their January 2008 search marketing report highlighting user and spending trends. The report paints a very rosy picture for search in the next three / four years with spending doubling from $8 billion this year to a projected $16 billion in 2011. These types of growth numbers clearly indicate we’re still in the growth phase of the s-curve. Some questions I had when reading the report…

1) Where are we on the s-curve? Is the industry still growing at a high rate?
We’re still growing at a tremendous rate.

Could the looming recession impact these numbers? You bet. But, it’s also likely MORE marketing dollars are poured into search. The savvy marketer will turn the screws on their campaigns and spend their marketing dollars on the most effective marketing channels that have a direct and measurable impact on their business’s bottom-line (i.e. search). eMarketer projects a slight decline in search’s share in marketer’s overall online marketing mix (see the chart below) but that could change. We could see an acceleration in the shifting of marketing dollars to better performing channels (i.e. search) which would lead to even greater growth for the industry (i.e. Google).

2) Where is the growth coming from?
eMarketer projects steady growth in paid search, paid inclusion, contextual advertising, and SEO through 2012.

While I do believe paid search and contextual advertising will grow, I do not believe paid inclusion and SEO will (at least not at the rates projected by eMarketer). Three reasons:

• SEO Expertise & The IT Department: SEO awareness is increasing and has finally caught the attention of the Technology department. Expertise in SEO is now becoming a requirement for developers and as that knowledge and expertise grows, there will be less of a need for SEO consultants. Websites will be built properly from the start. Further, as technology organizations within large companies become more integrated into the business and share responsibility for business metrics they’ll be forced to pay closer attention to SEO. Many large organizations still view technology teams as ’supporters’ rather than ‘drivers’ and that is starting to change. More on this evolution and how it effects search in the coming weeks.
• Search Engine Transparency: As the search engines increase transparency to their processes and produce tools to facilitate SEO, the need for outsourced technology lessens.
• The Future of Search: The evolution of the semantic web could dramatically change the presentation of search results and the source of content by 2012. Another post that’s on the back burner.

3) What is the latest search market share?
No surprises here. Note, the percentages exceed 100% because people visit more than one search engine.

4) Is consumer usage growing and where?
Absolutely. This is really a no-brainer as two things are working in search’s favor. One, search is a fundamental tool for Internet usage. As Internet usage grows so does search. The pie increases organically every year. Two, as these new Internet users increase their time on the net they inevitably find a need for search engines.

5) Tell me something I didn’t already know!
The upstream data provided by Hitwise was certainly eye-opening. A tremendous percentage of Yahoo and MSN’s search traffic originate from their portals and mail services. If they ever tie the knot and try consolidating those services, mucking things up along the way, Google could be in for an even bigger piece of the pie.

(Chart & data credit: eMarketer)

Comments (0) | Categories: Search Marketing Metrics | Tags: metrics, search engine market share, search engine spend, search engine usage    |  Popularity: 23% [?]